Premature birth remains a major public health issue due to its continuing high incidence combined with the frequency of developmental impairments among surviving infants. In order to improve developmental outcomes in very preterm infants (VPT, born prior to 32 weeks' gestation), accurate and early identification of the alterations in brain development that lead to subsequent developmental impairment is required. The candidate for this K02 Independent Scientist Award is a uniquely trained pediatric neurologist who seeks to define the relationship between cerebral connectivity during infancy and childhood developmental outcomes. This project will involve a valuable, well-characterized VPT cohort (N=128) which includes high-risk children selected due to the presence of brain injury during the neonatal period. These children have undergone clinical characterization and MRI during the neonatal period and developmental assessment at age 2 years. They will now undergo standardized assessment of developmental outcomes at age 5 years. The central hypotheses are that VPT children will display a persisting range of motor and language difficulties, and that affected VPT children will display abnormalities on imaging measures obtained during infancy that predict childhood performance, both at the group and individual level, in these distinct and critical developmental domains. We propose to investigate these postulates by: 1) comprehensively evaluating childhood motor and language performance; 2) defining the relationship between alterations in domain-specific neonatal functional and structural connectivity and childhood motor and language outcomes; and 3) investigating differences in brain- wide cerebral connectivity in impaired VPT children using machine learning, pattern classification approach. The result will be characterization of the alterations in cerebral connectivity that underlie developmental disabilities in VPT children. In addition, we seek to develop highly valuable neuroimaging biomarkers that predict future motor and language impairment, enabling the provision of targeted care tailored to the needs of each individual infant. Finally, these tools will contribute to future, large-scale trials evaluating interventions designed to protect the developing brain. In pursuit of this goal, a focused career development program will complement the candidate's background in engineering and magnetic resonance methodology with training in developmental assessments and innovative image analysis approaches using applied mathematics, providing skills necessary to perform accurate longitudinal clinical research studies and develop individual prediction models. The supported activities will facilitate an NIH R-mechanism application centered upon innovative investigation of individual prediction models for developmental trajectories in VPT children during school-age, including repeat neuroimaging assessments during childhood. The culmination of these efforts will be transformative investigation that fosters improved outcomes for this high-risk population.